Publications
2022 |
Žlajpah, Leon; Petrič, Tadej Kinematic calibration for collaborative robots on a mobile platform using motion capture system Journal Article Robotics and Computer-Integrated Manufacturing, 79 , pp. 102446, 2022, ISSN: 0736-5845. @article{zlajpah2022, title = {Kinematic calibration for collaborative robots on a mobile platform using motion capture system}, author = {Leon Žlajpah and Tadej Petrič}, url = {https://www.sciencedirect.com/science/article/pii/S0736584522001296}, doi = {https://doi.org/10.1016/j.rcim.2022.102446}, issn = {0736-5845}, year = {2022}, date = {2022-09-01}, journal = {Robotics and Computer-Integrated Manufacturing}, volume = {79}, pages = {102446}, abstract = {For modern robotic applications that go beyond the typical industrial environment, absolute accuracy is one of the key properties that make this possible. There are several approaches in the literature to improve robot accuracy for a typical industrial robot mounted on a fixed frame. In contrast, there is no method to improve robot accuracy when the robot is mounted on a mobile base, which is typical for collaborative robots. Therefore, in this work, we proposed and analyzed two approaches to improve the absolute accuracy of the robot mounted on a mobile platform using an optical measurement system. The first approach is based on geometric operations used to calculate the rotation axes of each joint. This approach identifies all rotational axes, which allows the calculation of the Denavit–Hartenberg (DH) parameters and thus the complete kinematic model, including the position and orientation errors of the robot end-effector and the robot base. The second approach to parameter estimation is based on optimization using a set of joint positions and end-effector poses to find the optimal DH parameters. Since the robot is mounted on a mobile base that is not fixed, an optical measurement system was used to dynamically and simultaneously measure the position of the robot base and the end-effector. The performance of the two proposed methods was analyzed and validated on a 7-DoF Franka Emika Panda robot mounted on a mobile platform PAL Tiago-base. The results show a significant improvement in absolute accuracy for both proposed approaches. By using the proposed approach with the optical measurement system, we can easily automate the estimation of robot kinematic parameters with the aim of improving absolute accuracy, especially in applications that require high positioning accuracy.}, keywords = {}, pubstate = {published}, tppubtype = {article} } For modern robotic applications that go beyond the typical industrial environment, absolute accuracy is one of the key properties that make this possible. There are several approaches in the literature to improve robot accuracy for a typical industrial robot mounted on a fixed frame. In contrast, there is no method to improve robot accuracy when the robot is mounted on a mobile base, which is typical for collaborative robots. Therefore, in this work, we proposed and analyzed two approaches to improve the absolute accuracy of the robot mounted on a mobile platform using an optical measurement system. The first approach is based on geometric operations used to calculate the rotation axes of each joint. This approach identifies all rotational axes, which allows the calculation of the Denavit–Hartenberg (DH) parameters and thus the complete kinematic model, including the position and orientation errors of the robot end-effector and the robot base. The second approach to parameter estimation is based on optimization using a set of joint positions and end-effector poses to find the optimal DH parameters. Since the robot is mounted on a mobile base that is not fixed, an optical measurement system was used to dynamically and simultaneously measure the position of the robot base and the end-effector. The performance of the two proposed methods was analyzed and validated on a 7-DoF Franka Emika Panda robot mounted on a mobile platform PAL Tiago-base. The results show a significant improvement in absolute accuracy for both proposed approaches. By using the proposed approach with the optical measurement system, we can easily automate the estimation of robot kinematic parameters with the aim of improving absolute accuracy, especially in applications that require high positioning accuracy. |
Gams, Andrej; č, Tadej Petri; Nemec, Bojan; Ude, Aleš Manipulation Learning on Humanoid Robots Journal Article Current Robotics Reports, 3 (3), pp. 97-109, 2022, ISSN: 2662-4087. @article{Gams2022, title = {Manipulation Learning on Humanoid Robots}, author = {Andrej Gams and Tadej Petri{č} and Bojan Nemec and Aleš Ude}, url = {https://doi.org/10.1007/s43154-022-00082-9}, doi = {10.1007/s43154-022-00082-9}, issn = {2662-4087}, year = {2022}, date = {2022-09-01}, journal = {Current Robotics Reports}, volume = {3}, number = {3}, pages = {97-109}, abstract = {The ability to autonomously manipulate the physical world is the key capability needed to fulfill the potential of cognitive robots. Humanoid robots, which offer very rich sensorimotor capabilities, have made giant leaps in their manipulation capabilities in recent years. Due to their similarity to humans, the progress can be partially attributed to the learning by demonstration paradigm. Supplemented by the autonomous learning methods to refine the demonstrated manipulation actions, humanoid robots can effectively learn new manipulation skills. In this paper we present continuous effort by our research group to advance the manipulation capabilities of humanoid robots and bring them to autonomously act in an unstructured world.}, keywords = {}, pubstate = {published}, tppubtype = {article} } The ability to autonomously manipulate the physical world is the key capability needed to fulfill the potential of cognitive robots. Humanoid robots, which offer very rich sensorimotor capabilities, have made giant leaps in their manipulation capabilities in recent years. Due to their similarity to humans, the progress can be partially attributed to the learning by demonstration paradigm. Supplemented by the autonomous learning methods to refine the demonstrated manipulation actions, humanoid robots can effectively learn new manipulation skills. In this paper we present continuous effort by our research group to advance the manipulation capabilities of humanoid robots and bring them to autonomously act in an unstructured world. |
Miskovic, Luka; Deman, Miha; Petric, Tadej Pneumatic quasi-passive variable stiffness mechanism for energy storage applications Journal Article IEEE Robotics and Automation Letters, pp. 1-1, 2022. @article{9674781, title = {Pneumatic quasi-passive variable stiffness mechanism for energy storage applications}, author = {Luka Miskovic and Miha Deman and Tadej Petric}, doi = {10.1109/LRA.2022.3141211}, year = {2022}, date = {2022-01-01}, journal = {IEEE Robotics and Automation Letters}, pages = {1-1}, keywords = {}, pubstate = {published}, tppubtype = {article} } |
Leskovar, Rebeka Kropivšek; Čamernik, Jernej; Petrič, Tadej Leader-Follower Dynamics in Complex Obstacle Avoidance Task Journal Article arXiv preprint arXiv:2207.04791, 2022. @article{leskovar2022leader, title = {Leader-Follower Dynamics in Complex Obstacle Avoidance Task}, author = {Rebeka Kropivšek Leskovar and Jernej Čamernik and Tadej Petrič}, url = {https://arxiv.org/abs/2207.04791}, doi = {https://doi.org/10.48550/arXiv.2207.04791}, year = {2022}, date = {2022-01-01}, journal = {arXiv preprint arXiv:2207.04791}, keywords = {}, pubstate = {published}, tppubtype = {article} } |
Leskovar, Rebeka Kropivšek; Petrič;, Tadej Increased Complexity of a Human-Robot Collaborative Task May Increase the Need for a Socially Competent Robot Inproceedings 2022 IEEE International Conference on Advanced Robotics and Its Social Impacts (ARSO), pp. 1-6, 2022. @inproceedings{9802968, title = {Increased Complexity of a Human-Robot Collaborative Task May Increase the Need for a Socially Competent Robot}, author = {Rebeka Kropivšek Leskovar and Tadej Petrič;}, doi = {10.1109/ARSO54254.2022.9802968}, year = {2022}, date = {2022-01-01}, booktitle = {2022 IEEE International Conference on Advanced Robotics and Its Social Impacts (ARSO)}, pages = {1-6}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } |
Brecelj, Tilen; Petrič, Tadej Zero Moment Line—Universal Stability Parameter for Multi-Contact Systems in Three Dimensions Journal Article Sensors, 22 (15), 2022, ISSN: 1424-8220. @article{Brecelj2022, title = {Zero Moment Line—Universal Stability Parameter for Multi-Contact Systems in Three Dimensions}, author = {Tilen Brecelj and Tadej Petrič}, url = {https://www.mdpi.com/1424-8220/22/15/5656}, doi = {10.3390/s22155656}, issn = {1424-8220}, year = {2022}, date = {2022-01-01}, journal = {Sensors}, volume = {22}, number = {15}, abstract = {The widely used stability parameter, the zero moment point (ZMP), which is usually defined on the ground, is redefined, in this paper, in two different ways to acquire a more general form that allows its application to systems that are not supported only on the ground, and therefore, their support polygon does not extend only on the floor. This way it allows to determine the stability of humanoid and other floating-based robots that are interacting with the environment at arbitrary heights. In the first redefinition, the ZMP is represented as a line containing all possible ZMPs, called the zero moment line (ZML), while in the second redefinition, the ZMP is represented as the ZMP angle, i.e., the angle between the ZML and the vertical line, passing through the center of mass (COM) of the investigated system. The first redefinition is useful in situations when the external forces and their acting locations are known, while the second redefinition can be applied in situations when the COM of the system under study is known and can be tracked. The first redefinition of the ZMP is also applied to two different measurements performed with two force plates, two force sensors, and the Optitrack system. In the first measurement, a subject stands up from a bench and sits down while being pulled by its hands, while in the second measurement, two subjects stand still, hold on to two double handles, and lean backward. In both cases, the stability of the subjects involved in the measurements is investigated and discussed.}, keywords = {}, pubstate = {published}, tppubtype = {article} } The widely used stability parameter, the zero moment point (ZMP), which is usually defined on the ground, is redefined, in this paper, in two different ways to acquire a more general form that allows its application to systems that are not supported only on the ground, and therefore, their support polygon does not extend only on the floor. This way it allows to determine the stability of humanoid and other floating-based robots that are interacting with the environment at arbitrary heights. In the first redefinition, the ZMP is represented as a line containing all possible ZMPs, called the zero moment line (ZML), while in the second redefinition, the ZMP is represented as the ZMP angle, i.e., the angle between the ZML and the vertical line, passing through the center of mass (COM) of the investigated system. The first redefinition is useful in situations when the external forces and their acting locations are known, while the second redefinition can be applied in situations when the COM of the system under study is known and can be tracked. The first redefinition of the ZMP is also applied to two different measurements performed with two force plates, two force sensors, and the Optitrack system. In the first measurement, a subject stands up from a bench and sits down while being pulled by its hands, while in the second measurement, two subjects stand still, hold on to two double handles, and lean backward. In both cases, the stability of the subjects involved in the measurements is investigated and discussed. |
2021 |
Leskovar, Rebeka Kropivšek; Čamernik, Jernej; Petrič, Tadej Leader–Follower Role Allocation for Physical Collaboration in Human Dyads Journal Article Applied Sciences, 11 (19), 2021, ISSN: 2076-3417. @article{Leskovar2021, title = {Leader–Follower Role Allocation for Physical Collaboration in Human Dyads}, author = {Rebeka Kropivšek Leskovar and Jernej Čamernik and Tadej Petrič}, url = {https://www.mdpi.com/2076-3417/11/19/8928}, doi = {10.3390/app11198928}, issn = {2076-3417}, year = {2021}, date = {2021-09-25}, journal = {Applied Sciences}, volume = {11}, number = {19}, abstract = {People often find themselves in situations where collaboration with others is necessary to accomplish a particular task. In such cases, a leader–follower relationship is established to coordinate a plan to achieve a common goal. This is usually accomplished through verbal communication. However, what happens when verbal communication is not possible? In this study, we observe the dynamics of a leader–follower relationship in human dyads during collaborative tasks where there is no verbal communication between partners. Using two robotic arms, we designed a collaborative experimental task in which subjects perform the task individually or coupled together through a virtual model. The results show that human partners fall into the leader–follower dynamics even when they cannot communicate verbally. We demonstrate this in two steps. First, we study how each subject in a collaboration influences task performance, and second, we evaluate whether both partners influence it equally or not using our proposed sorting method to objectively identify a leader. We also study the leader–follower dynamics by analysing the task performance of partners during their individual sessions to predict the role distribution in a dyad. Based on the results of our prediction method, we conclude that the higher-performing individual performance will assume the role of a leader in collaboration.}, keywords = {}, pubstate = {published}, tppubtype = {article} } People often find themselves in situations where collaboration with others is necessary to accomplish a particular task. In such cases, a leader–follower relationship is established to coordinate a plan to achieve a common goal. This is usually accomplished through verbal communication. However, what happens when verbal communication is not possible? In this study, we observe the dynamics of a leader–follower relationship in human dyads during collaborative tasks where there is no verbal communication between partners. Using two robotic arms, we designed a collaborative experimental task in which subjects perform the task individually or coupled together through a virtual model. The results show that human partners fall into the leader–follower dynamics even when they cannot communicate verbally. We demonstrate this in two steps. First, we study how each subject in a collaboration influences task performance, and second, we evaluate whether both partners influence it equally or not using our proposed sorting method to objectively identify a leader. We also study the leader–follower dynamics by analysing the task performance of partners during their individual sessions to predict the role distribution in a dyad. Based on the results of our prediction method, we conclude that the higher-performing individual performance will assume the role of a leader in collaboration. |
Simonič, Mihael; Petrič, Tadej; Ude, Aleš; Nemec, Bojan Analysis of Methods for Incremental Policy Refinement by Kinesthetic Guidance Journal Article Journal of Intelligent & Robotic Systems, 102 (1), pp. 5, 2021, ISSN: 0921-0296. @article{Simonic2021, title = {Analysis of Methods for Incremental Policy Refinement by Kinesthetic Guidance}, author = {Mihael Simonič and Tadej Petrič and Aleš Ude and Bojan Nemec}, url = {https://link.springer.com/10.1007/s10846-021-01328-y}, doi = {10.1007/s10846-021-01328-y}, issn = {0921-0296}, year = {2021}, date = {2021-05-01}, journal = {Journal of Intelligent & Robotic Systems}, volume = {102}, number = {1}, pages = {5}, abstract = {Traditional robot programming is often not feasible in small-batch production, as it is time-consuming, inefficient, and expensive. To shorten the time necessary to deploy robot tasks, we need appropriate tools to enable efficient reuse of existing robot control policies. Incremental Learning from Demonstration (iLfD) and reversible Dynamic Movement Primitives (DMP) provide a framework for efficient policy demonstration and adaptation. In this paper, we extend our previously proposed framework with improvements that provide better performance and lower the algorithm's computational burden. Further, we analyse the learning stability and evaluate the proposed framework with a comprehensive user study. The proposed methods have been evaluated on two popular collaborative robots, Franka Emika Panda and Universal Robot UR10.}, keywords = {}, pubstate = {published}, tppubtype = {article} } Traditional robot programming is often not feasible in small-batch production, as it is time-consuming, inefficient, and expensive. To shorten the time necessary to deploy robot tasks, we need appropriate tools to enable efficient reuse of existing robot control policies. Incremental Learning from Demonstration (iLfD) and reversible Dynamic Movement Primitives (DMP) provide a framework for efficient policy demonstration and adaptation. In this paper, we extend our previously proposed framework with improvements that provide better performance and lower the algorithm's computational burden. Further, we analyse the learning stability and evaluate the proposed framework with a comprehensive user study. The proposed methods have been evaluated on two popular collaborative robots, Franka Emika Panda and Universal Robot UR10. |
Knezevic, Nikola; Lukic, Branko; Jovanovic, Kosta; Zlajpah, Leon; Petric, Tadej End-effector Cartesian stiffness shaping - sequential least squares programming approach Journal Article Serbian Journal of Electrical Engineering, 18 (1), pp. 1–14, 2021, ISSN: 1451-4869. @article{Knezevic2021, title = {End-effector Cartesian stiffness shaping - sequential least squares programming approach}, author = {Nikola Knezevic and Branko Lukic and Kosta Jovanovic and Leon Zlajpah and Tadej Petric}, url = {http://www.doiserbia.nb.rs/Article.aspx?ID=1451-48692101001K http://cobotat.ijs.si/wp-content/uploads/2021/04/Knezevic-et-al._2021_End-effector-Cartesian-stiffness-shaping-sequential-least-squares-programming-approach_Serbian-Journal-of-Electri.pdf}, doi = {10.2298/SJEE2101001K}, issn = {1451-4869}, year = {2021}, date = {2021-01-01}, journal = {Serbian Journal of Electrical Engineering}, volume = {18}, number = {1}, pages = {1--14}, abstract = {Control of robot end-effector (EE) Cartesian stiffness matrix (or the whole mechanical impedance) is still a challenging open issue in physical humanrobot interaction (pHRI). This paper presents an optimization approach for shaping the robot EE Cartesian stiffness. This research targets collaborative robots with intrinsic compliance – serial elastic actuators (SEAs). Although robots with SEAs have constant joint stiffness, task redundancy (null-space) for a specific task could be used for robot reconfiguration and shaping the stiffness matrix while still keeping the EE position unchanged. The method proposed in this paper to investigate null-space reconfiguration's influence on Cartesian robot stiffness is based on the Sequential Least Squares Programming (SLSQP) algorithm, which presents an expansion of the quadratic programming algorithm for nonlinear functions with constraints. The method is tested in simulations for 4 DOF planar robot. Results are presented for control of the EE Cartesian stiffness initially along one axis, and then control of stiffness along both planar axis – shaping the main diagonal of the EE stiffness matrix.}, keywords = {}, pubstate = {published}, tppubtype = {article} } Control of robot end-effector (EE) Cartesian stiffness matrix (or the whole mechanical impedance) is still a challenging open issue in physical humanrobot interaction (pHRI). This paper presents an optimization approach for shaping the robot EE Cartesian stiffness. This research targets collaborative robots with intrinsic compliance – serial elastic actuators (SEAs). Although robots with SEAs have constant joint stiffness, task redundancy (null-space) for a specific task could be used for robot reconfiguration and shaping the stiffness matrix while still keeping the EE position unchanged. The method proposed in this paper to investigate null-space reconfiguration's influence on Cartesian robot stiffness is based on the Sequential Least Squares Programming (SLSQP) algorithm, which presents an expansion of the quadratic programming algorithm for nonlinear functions with constraints. The method is tested in simulations for 4 DOF planar robot. Results are presented for control of the EE Cartesian stiffness initially along one axis, and then control of stiffness along both planar axis – shaping the main diagonal of the EE stiffness matrix. |
Brecelj, Tilen; Petric, Tadej Angular Dependency of the Zero Moment Point Inproceedings Zeghloul, Said; Laribi, Med Amine; Sandoval, Juan (Ed.): Advances in Service and Industrial Robotics. RAAD 2021, pp. 135–144, Springer International Publishing, Cham, 2021, ISBN: 978-3-030-75259-0. @inproceedings{10.1007/978-3-030-75259-0_15, title = {Angular Dependency of the Zero Moment Point}, author = {Tilen Brecelj and Tadej Petric}, editor = {Said Zeghloul and Med Amine Laribi and Juan Sandoval}, doi = {https://doi.org/10.1007/978-3-030-75259-0_15}, isbn = {978-3-030-75259-0}, year = {2021}, date = {2021-01-01}, booktitle = {Advances in Service and Industrial Robotics. RAAD 2021}, pages = {135--144}, publisher = {Springer International Publishing}, address = {Cham}, series = {Mechanisms and Machine Science}, abstract = {In this paper, the standard definition of the zero moment point as a location on the ground is extended to its angular definition around the center of mass of a humanoid robot. This new definition provides a more general way of expressing this stability parameter and enables its wider use, as this way it can be located anywhere on a specific line passing through the system's center of mass. The angular expression of the zero moment point is examined and compared with the zero moment point of the linear inverted pendulum model, which can be obtained with some simplifications of its angular definition. For a better understanding of the angular dependence of the zero moment point with respect to the accelerations of the humanoid's center of mass in the horizontal and vertical directions, some computer simulations are presented. Finally, a real scenario of humanoid postural stability is presented and discussed, in which the advantages of the angular definition of the zero moment point can be seen.}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } In this paper, the standard definition of the zero moment point as a location on the ground is extended to its angular definition around the center of mass of a humanoid robot. This new definition provides a more general way of expressing this stability parameter and enables its wider use, as this way it can be located anywhere on a specific line passing through the system's center of mass. The angular expression of the zero moment point is examined and compared with the zero moment point of the linear inverted pendulum model, which can be obtained with some simplifications of its angular definition. For a better understanding of the angular dependence of the zero moment point with respect to the accelerations of the humanoid's center of mass in the horizontal and vertical directions, some computer simulations are presented. Finally, a real scenario of humanoid postural stability is presented and discussed, in which the advantages of the angular definition of the zero moment point can be seen. |
Leskovar, Rebeka Kropivšek; Petrič, Tadej Humans Prefer Collaborating with a Robot Who Leads in a Physical Human-Robot Collaboration Scenario Inproceedings 2021 20th International Conference on Advanced Robotics (ICAR), pp. 935-941, 2021. @inproceedings{9659365, title = {Humans Prefer Collaborating with a Robot Who Leads in a Physical Human-Robot Collaboration Scenario}, author = {Rebeka Kropivšek Leskovar and Tadej Petrič}, doi = {10.1109/ICAR53236.2021.9659365}, year = {2021}, date = {2021-01-01}, booktitle = {2021 20th International Conference on Advanced Robotics (ICAR)}, pages = {935-941}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } |
Mišković, Luka; Dežman, Miha; Petrić, Tadej Modular quasi-passive mechanism for energy storage applications: towards lightweight high-performance exoskeleton Inproceedings 2021 20th International Conference on Advanced Robotics (ICAR), pp. 588-593, 2021. @inproceedings{9659353, title = {Modular quasi-passive mechanism for energy storage applications: towards lightweight high-performance exoskeleton}, author = {Luka Mišković and Miha Dežman and Tadej Petrić}, doi = {10.1109/ICAR53236.2021.9659353}, year = {2021}, date = {2021-01-01}, booktitle = {2021 20th International Conference on Advanced Robotics (ICAR)}, pages = {588-593}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } |
2020 |
Rebeka, Kropivšek Leskovar ; Petrič, Tadej Turing test of motor ability perception in physical collaboration between a human and an intelligent robot agent Inproceedings Žemva Andrej, Trost Andrej (Ed.): Proceedings of the Twenty-ninth International Electrotechnical and Computer Science Conference ERK 2020, 2020, ISBN: 2591-0442, 29. @inproceedings{kropivsek2020, title = {Turing test of motor ability perception in physical collaboration between a human and an intelligent robot agent}, author = {Rebeka, Kropivšek Leskovar and Tadej Petrič}, editor = {Žemva, Andrej, Trost, Andrej}, url = {http://cobotat.ijs.si/wp-content/uploads/2021/04/ERK2020.pdf}, isbn = {2591-0442, 29}, year = {2020}, date = {2020-09-21}, booktitle = {Proceedings of the Twenty-ninth International Electrotechnical and Computer Science Conference ERK 2020}, volume = {ERK 2020}, abstract = {In this paper we propose a novel robot control method for human-robot collaboration tasks that takes into account the leader-follower relationship found in human interactions. Taking into account the leader-follower dynamics, learnt during a study on human-human collaboration, the control method replicates human behaviour when performing collaborative tasks. The performance of the proposed control method was evaluated using a 2D reaching task where we compared task performance between individual tasks, tasks in collaboration with a human and tasks in collaboration with a robot. The subjects in the evaluation were asked to grade their perceived task load for each experiment as well as specify if they thought they performed the task alone, with a robot or with a human partner as a Turing test to determine whether the subjects were able to distinct between a robot and a human partner. The results of the evaluation showed, that the robot control method is capable of replicating human behavior to benefit overall task performance of the subject in collaboration, however it is not capable of replicating this behaviour to the degree that the subject in collaboration would not be able distinct whether they were collaborating with a robot or a human partner. }, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } In this paper we propose a novel robot control method for human-robot collaboration tasks that takes into account the leader-follower relationship found in human interactions. Taking into account the leader-follower dynamics, learnt during a study on human-human collaboration, the control method replicates human behaviour when performing collaborative tasks. The performance of the proposed control method was evaluated using a 2D reaching task where we compared task performance between individual tasks, tasks in collaboration with a human and tasks in collaboration with a robot. The subjects in the evaluation were asked to grade their perceived task load for each experiment as well as specify if they thought they performed the task alone, with a robot or with a human partner as a Turing test to determine whether the subjects were able to distinct between a robot and a human partner. The results of the evaluation showed, that the robot control method is capable of replicating human behavior to benefit overall task performance of the subject in collaboration, however it is not capable of replicating this behaviour to the degree that the subject in collaboration would not be able distinct whether they were collaborating with a robot or a human partner. |
Petrič, Tadej; Jamšek, Marko; Babič, Jan Exoskeleton Control Based on Network of Stable Heteroclinic Channels (SHC) Combined with Gaussian Mixture Models (GMM) Inproceedings Lenarčič, Jadran; Siciliano, Bruno (Ed.): pp. 341–348, Springer International Publishing, Cham, 2020, ISBN: 978-3-030-50975-0. @inproceedings{petric2020, title = {Exoskeleton Control Based on Network of Stable Heteroclinic Channels (SHC) Combined with Gaussian Mixture Models (GMM)}, author = {Tadej Petrič and Marko Jamšek and Jan Babič}, editor = {Jadran Lenarčič and Bruno Siciliano}, url = {https://doi.org/10.1007/978-3-030-50975-0_42}, isbn = {978-3-030-50975-0}, year = {2020}, date = {2020-07-18}, pages = {341--348}, publisher = {Springer International Publishing}, address = {Cham}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } |
Jamšek, Marko; Petrič, Tadej; Babič, Jan Gaussian Mixture Models for Control of Quasi-Passive Spinal Exoskeletons Journal Article Sensors, 20 (9), 2020, ISSN: 1424-8220. @article{Jamsek2020, title = {Gaussian Mixture Models for Control of Quasi-Passive Spinal Exoskeletons}, author = {Marko Jamšek and Tadej Petrič and Jan Babič}, url = {https://www.mdpi.com/1424-8220/20/9/2705}, doi = {10.3390/s20092705}, issn = {1424-8220}, year = {2020}, date = {2020-01-01}, journal = {Sensors}, volume = {20}, number = {9}, keywords = {}, pubstate = {published}, tppubtype = {article} } |
Žlajpah, Leon; Petrič, Tadej Generation of Smooth Cartesian Paths Using Radial Basis Functions Inproceedings Zeghloul, Said; Laribi, Med Amine; Arevalo, Juan Sebastian Sandoval (Ed.): Advances in Service and Industrial Robotics, pp. 171–180, Springer International Publishing, Cham, 2020, ISBN: 978-3-030-48989-2. @inproceedings{10.1007/978-3-030-48989-2_19, title = {Generation of Smooth Cartesian Paths Using Radial Basis Functions}, author = {Leon Žlajpah and Tadej Petrič}, editor = {Said Zeghloul and Med Amine Laribi and Juan Sebastian Sandoval Arevalo}, isbn = {978-3-030-48989-2}, year = {2020}, date = {2020-01-01}, booktitle = {Advances in Service and Industrial Robotics}, pages = {171--180}, publisher = {Springer International Publishing}, address = {Cham}, abstract = {In this paper, we consider the problem of generating smooth Cartesian paths for robots passing through a sequence of waypoints. For interpolation between waypoints we propose to use radial basis functions (RBF). First, we describe RBF based on Gaussian kernel functions and how the weights are calculated. The path generation considers also boundary conditions for velocity and accelerations. Then we present how RBF parameters influence the shape of the generated path. The proposed RBF method is compared with paths generated by a spline and linear interpolation. The results demonstrate the advantages of the proposed method, which is offering a good alternative to generate smooth Cartesian paths.}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } In this paper, we consider the problem of generating smooth Cartesian paths for robots passing through a sequence of waypoints. For interpolation between waypoints we propose to use radial basis functions (RBF). First, we describe RBF based on Gaussian kernel functions and how the weights are calculated. The path generation considers also boundary conditions for velocity and accelerations. Then we present how RBF parameters influence the shape of the generated path. The proposed RBF method is compared with paths generated by a spline and linear interpolation. The results demonstrate the advantages of the proposed method, which is offering a good alternative to generate smooth Cartesian paths. |
Petrič, Tadej; Žlajpah, Leon Combining Virtual and Physical Guides for Autonomous In-Contact Path Adaptation Inproceedings Zeghloul, Said; Laribi, Med Amine; Arevalo, Juan Sebastian Sandoval (Ed.): Advances in Service and Industrial Robotics, pp. 181–189, Springer International Publishing, Cham, 2020, ISBN: 978-3-030-48989-2. @inproceedings{10.1007/978-3-030-48989-2_20, title = {Combining Virtual and Physical Guides for Autonomous In-Contact Path Adaptation}, author = {Tadej Petrič and Leon Žlajpah}, editor = {Said Zeghloul and Med Amine Laribi and Juan Sebastian Sandoval Arevalo}, url = {http://cobotat.ijs.si/wp-content/uploads/2021/04/Raad2020.pdf}, isbn = {978-3-030-48989-2}, year = {2020}, date = {2020-01-01}, booktitle = {Advances in Service and Industrial Robotics}, pages = {181--189}, publisher = {Springer International Publishing}, address = {Cham}, abstract = {Several approaches exist for learning and control of robot behaviors in physical human-robot interaction (PHRI) scenarios. One of these is the approach based on virtual guides which actively helps to guide the user. Such a system enables guiding users towards preferred movement directions or prevents them to enter into a prohibited zone. Despite being shown that such a framework works well in physical contact with humans, the efficient interaction with the environment is still limited. Within the virtual guide framework, the environment is considered as a physical guide, for example, a table is a plane that prevents the robot to penetrate through. To mitigate these limits we introduce and evaluate the means of autonomous path adaptation through interaction with physical guides, which essentially means merging virtual and physical guides. The virtual guide framework was extended by introducing an algorithm which partially modifies the virtual guides online. The path updates are now based on the interactive force measurements and essentially improves the virtual guides to match them with the actual physical guides.}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } Several approaches exist for learning and control of robot behaviors in physical human-robot interaction (PHRI) scenarios. One of these is the approach based on virtual guides which actively helps to guide the user. Such a system enables guiding users towards preferred movement directions or prevents them to enter into a prohibited zone. Despite being shown that such a framework works well in physical contact with humans, the efficient interaction with the environment is still limited. Within the virtual guide framework, the environment is considered as a physical guide, for example, a table is a plane that prevents the robot to penetrate through. To mitigate these limits we introduce and evaluate the means of autonomous path adaptation through interaction with physical guides, which essentially means merging virtual and physical guides. The virtual guide framework was extended by introducing an algorithm which partially modifies the virtual guides online. The path updates are now based on the interactive force measurements and essentially improves the virtual guides to match them with the actual physical guides. |
Leskovar, Rebeka Kropivšek; Čamernik, Jernej; Petrič, Tadej Dyadic Human-Human Interactions in Reaching Tasks: Fitts' Law for Two Inproceedings Zeghloul, Said; Laribi, Med Amine; Arevalo, Juan Sebastian Sandoval (Ed.): Advances in Service and Industrial Robotics, pp. 199–207, Springer International Publishing, Cham, 2020, ISBN: 978-3-030-48989-2. @inproceedings{10.1007/978-3-030-48989-2_22, title = {Dyadic Human-Human Interactions in Reaching Tasks: Fitts' Law for Two}, author = {Rebeka Kropivšek Leskovar and Jernej Čamernik and Tadej Petrič}, editor = {Said Zeghloul and Med Amine Laribi and Juan Sebastian Sandoval Arevalo}, url = {http://cobotat.ijs.si/wp-content/uploads/2021/04/RAAD2020_Rkl_Final.pdf}, isbn = {978-3-030-48989-2}, year = {2020}, date = {2020-01-01}, booktitle = {Advances in Service and Industrial Robotics}, pages = {199--207}, publisher = {Springer International Publishing}, address = {Cham}, abstract = {In this paper we examine physical collaboration between two individuals using a dual-arm robot as a haptic interface. First, we design a haptic controller based on a virtual dynamic model of the robot arms. Then, we analyse dyadic human-human collaboration with a reaching task on a 2D plane, where the distance and size of the target changed randomly from a pool of nine reachable positions and sizes. Each subject performed the task individually and linked through the guided robot arms with a virtual model to perform the same task in collaboration. We evaluated both, individual and collaborative performances, based on Fitts' law, which describes the relation between the speed of motion and its accuracy. The results show that the Fitts' law applies to both, individual and collaborative tasks, with their performance improving when in collaboration.}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } In this paper we examine physical collaboration between two individuals using a dual-arm robot as a haptic interface. First, we design a haptic controller based on a virtual dynamic model of the robot arms. Then, we analyse dyadic human-human collaboration with a reaching task on a 2D plane, where the distance and size of the target changed randomly from a pool of nine reachable positions and sizes. Each subject performed the task individually and linked through the guided robot arms with a virtual model to perform the same task in collaboration. We evaluated both, individual and collaborative performances, based on Fitts' law, which describes the relation between the speed of motion and its accuracy. The results show that the Fitts' law applies to both, individual and collaborative tasks, with their performance improving when in collaboration. |
Lukić, Branko; Jovanović, Kosta; Knežević, Nikola; Žlajpah, Leon; Petrič, Tadej Maximizing the End-Effector Cartesian Stiffness Range for Kinematic Redundant Robot with Compliance Inproceedings Zeghloul, Said; Laribi, Med Amine; Arevalo, Juan Sebastian Sandoval (Ed.): Advances in Service and Industrial Robotics, pp. 208–217, Springer International Publishing, Cham, 2020, ISBN: 978-3-030-48989-2. @inproceedings{10.1007/978-3-030-48989-2_23, title = {Maximizing the End-Effector Cartesian Stiffness Range for Kinematic Redundant Robot with Compliance}, author = {Branko Lukić and Kosta Jovanović and Nikola Knežević and Leon Žlajpah and Tadej Petrič}, editor = {Said Zeghloul and Med Amine Laribi and Juan Sebastian Sandoval Arevalo}, isbn = {978-3-030-48989-2}, year = {2020}, date = {2020-01-01}, booktitle = {Advances in Service and Industrial Robotics}, pages = {208--217}, publisher = {Springer International Publishing}, address = {Cham}, abstract = {Compliant robots with constant joint stiffness (Serial Elastic Actuators - SEA), on the contrary to ones with variable joint stiffness (Variable Stiffness Actuators -- VSA), have limited capabilities for modulating robot mechanical impedance in the interaction task. However, in the case of kinematic redundancy in specific tasks, robots can exploit the null space to adjust End-Effector (EE) Cartesian stiffness. Thus, prior knowledge of the task path or the operational workspace can be used to pre-compute joint stiffness that can enable maximal ratio between maximal and minimal stiffness of the robot's EE during the task execution, and therefore shape achievable EE stiffness to best fit the task execution. In that light, this paper elaborates on the preselection of joint stiffnesses which influences the achievable robot's Cartesian stiffness in a specific task. Besides optimizing the available operational EE stiffness, by pre-computed joint stiffness values, the robot will be able to adapt better to specific tasks and provide a better framework for safe and efficient physical human-robot interaction. The paper presents an approach to the selection of predefined joint stiffness values of the 7-DOFs KUKA LWR, where joint stiffness is achieved/emulated with torque feedback. In the simulation experiments, the approach is depicted in the preselection of two joint stiffness values within the prescribed range, while other joint stiffness is set constant.}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } Compliant robots with constant joint stiffness (Serial Elastic Actuators - SEA), on the contrary to ones with variable joint stiffness (Variable Stiffness Actuators -- VSA), have limited capabilities for modulating robot mechanical impedance in the interaction task. However, in the case of kinematic redundancy in specific tasks, robots can exploit the null space to adjust End-Effector (EE) Cartesian stiffness. Thus, prior knowledge of the task path or the operational workspace can be used to pre-compute joint stiffness that can enable maximal ratio between maximal and minimal stiffness of the robot's EE during the task execution, and therefore shape achievable EE stiffness to best fit the task execution. In that light, this paper elaborates on the preselection of joint stiffnesses which influences the achievable robot's Cartesian stiffness in a specific task. Besides optimizing the available operational EE stiffness, by pre-computed joint stiffness values, the robot will be able to adapt better to specific tasks and provide a better framework for safe and efficient physical human-robot interaction. The paper presents an approach to the selection of predefined joint stiffness values of the 7-DOFs KUKA LWR, where joint stiffness is achieved/emulated with torque feedback. In the simulation experiments, the approach is depicted in the preselection of two joint stiffness values within the prescribed range, while other joint stiffness is set constant. |
Petrič, Tadej Phase-Synchronized Learning of Periodic Compliant Movement Primitives (P-CMPs) Journal Article Frontiers in Neurorobotics, 14 , pp. 90, 2020, ISSN: 1662-5218. @article{10.3389/fnbot.2020.599889, title = {Phase-Synchronized Learning of Periodic Compliant Movement Primitives (P-CMPs)}, author = {Tadej Petrič}, url = {https://www.frontiersin.org/article/10.3389/fnbot.2020.599889 http://cobotat.ijs.si/wp-content/uploads/2021/04/Petric_2020_Phase-Synchronized-Learning-of-Periodic-Compliant-Movement-Primitives-P-CMPs_Frontiers-in-Neurorobotics.pdf}, doi = {10.3389/fnbot.2020.599889}, issn = {1662-5218}, year = {2020}, date = {2020-01-01}, journal = {Frontiers in Neurorobotics}, volume = {14}, pages = {90}, abstract = {Autonomous trajectory and torque profile synthesis through modulation and generalization require a database of motion with accompanying dynamics, which is typically difficult and time-consuming to obtain. Inspired by adaptive control strategies, this paper presents a novel method for learning and synthesizing Periodic Compliant Movement Primitives (P-CMPs). P-CMPs combine periodic trajectories encoded as Periodic Dynamic Movement Primitives (P-DMPs) with accompanying task-specific Periodic Torque Primitives (P-TPs). The state-of-the-art approach requires to learn TPs for each variation of the task, e.g., modulation of frequency. Comparatively, in this paper, we propose a novel P-TPs framework, which is both frequency and phase-dependent. Thereby, the executed P-CMPs can be easily modulated, and consequently, the learning rate can be improved. Moreover, both the kinematic and the dynamic profiles are parameterized, thus enabling the representation of skills using corresponding parameters. The proposed framework was evaluated on two robot systems, i.e., Kuka LWR-4 and Franka Emika Panda. The evaluation of the proposed approach on a Kuka LWR-4 robot performing a swinging motion and on Franka Emika Panda performing an exercise for elbow rehabilitation shows fast P-CTPs acquisition and accurate and compliant motion in real-world scenarios.}, keywords = {}, pubstate = {published}, tppubtype = {article} } Autonomous trajectory and torque profile synthesis through modulation and generalization require a database of motion with accompanying dynamics, which is typically difficult and time-consuming to obtain. Inspired by adaptive control strategies, this paper presents a novel method for learning and synthesizing Periodic Compliant Movement Primitives (P-CMPs). P-CMPs combine periodic trajectories encoded as Periodic Dynamic Movement Primitives (P-DMPs) with accompanying task-specific Periodic Torque Primitives (P-TPs). The state-of-the-art approach requires to learn TPs for each variation of the task, e.g., modulation of frequency. Comparatively, in this paper, we propose a novel P-TPs framework, which is both frequency and phase-dependent. Thereby, the executed P-CMPs can be easily modulated, and consequently, the learning rate can be improved. Moreover, both the kinematic and the dynamic profiles are parameterized, thus enabling the representation of skills using corresponding parameters. The proposed framework was evaluated on two robot systems, i.e., Kuka LWR-4 and Franka Emika Panda. The evaluation of the proposed approach on a Kuka LWR-4 robot performing a swinging motion and on Franka Emika Panda performing an exercise for elbow rehabilitation shows fast P-CTPs acquisition and accurate and compliant motion in real-world scenarios. |
2019 |
Žlajpah, L; Petrič, T Unified Virtual Guides Framework for Path Tracking Tasks Journal Article Robotica, pp. 1–17, 2019. @article{zlajpah2019_robotica, title = {Unified Virtual Guides Framework for Path Tracking Tasks}, author = {L. Žlajpah and T. Petrič}, doi = {10.1017/S0263574719000973}, year = {2019}, date = {2019-08-13}, urldate = {2019-08-13}, journal = {Robotica}, pages = {1–17}, publisher = {Cambridge University Press}, keywords = {}, pubstate = {published}, tppubtype = {article} } |
Babič, J; Petrič, T; Mombaur, K; Kingma, I; Bornmann, J; González-Vargas, J; Baltrusch, S; Šarabon, N; Houdijk, H SPEXOR: Design and development of passive spinal exoskeletal robot for low back pain prevention and vocational reintegration Journal Article SN Applied Sciences, 1 (3), pp. 262, 2019, ISSN: 2523-3971. @article{Babic2019, title = {SPEXOR: Design and development of passive spinal exoskeletal robot for low back pain prevention and vocational reintegration}, author = {J. Babič and T. Petrič and K. Mombaur and I. Kingma and J. Bornmann and J. González-Vargas and S. Baltrusch and N. Šarabon and H. Houdijk}, doi = {10.1007/s42452-019-0266-1}, issn = {2523-3971}, year = {2019}, date = {2019-02-23}, journal = {SN Applied Sciences}, volume = {1}, number = {3}, pages = {262}, abstract = {The objective of SPEXOR project is to address low back pain as one of the most appealing health problems of the modern society by creating a body of scientific and technological knowledge in the multidisciplinary areas of biomechanics, robotics, and computer science that will lead to technologies for low back pain prevention. In this paper we provide an overview of the current state-of-art of SPEXOR that the consortium achieved in the first twenty-four months of the project. After introducing the rationale, we describe the biomechanics of low back pain intervention, development of the musculoskeletal stress monitoring for assessment of neuromuscular trunk functions, modeling and optimization of the interaction of spinal exoskeleton with the human body, electromechanical design and development of the passive spinal exoskeleton and its control, and finally the end-user evaluation of the functional effects, usability and satisfaction.}, keywords = {}, pubstate = {published}, tppubtype = {article} } The objective of SPEXOR project is to address low back pain as one of the most appealing health problems of the modern society by creating a body of scientific and technological knowledge in the multidisciplinary areas of biomechanics, robotics, and computer science that will lead to technologies for low back pain prevention. In this paper we provide an overview of the current state-of-art of SPEXOR that the consortium achieved in the first twenty-four months of the project. After introducing the rationale, we describe the biomechanics of low back pain intervention, development of the musculoskeletal stress monitoring for assessment of neuromuscular trunk functions, modeling and optimization of the interaction of spinal exoskeleton with the human body, electromechanical design and development of the passive spinal exoskeleton and its control, and finally the end-user evaluation of the functional effects, usability and satisfaction. |
Petrič, T; Gams, A Task Space Torque Profile Adaptations for Dynamical Human-Robot Motion Transfer Inproceedings pp. 44–52, Springer International Publishing, Cham, 2019, ISBN: 978-3-030-00232-9. @inproceedings{Petric2019, title = {Task Space Torque Profile Adaptations for Dynamical Human-Robot Motion Transfer}, author = {T. Petrič and A. Gams}, isbn = {978-3-030-00232-9}, year = {2019}, date = {2019-01-01}, pages = {44--52}, publisher = {Springer International Publishing}, address = {Cham}, abstract = {Motion transfer from a human to a robot implies accurate tracking of desired, demonstrated trajectories. However, direct imitation of joint position trajectories might not result in similar behavior of the robot and the human, because they have different kinematic and dynamic properties, i.e., different embodiment. To avoid the correspondence problem, the demonstrated trajectories need to be somehow adapted. In this paper we go beyond simple imitation, but we show how the torque profiles that should execute the demonstrated position trajectories are being learned in a manner that preserves the correspondence. Thus, position trajectories are modified from the demonstration and, furthermore, the robot executes the motion that preserves correspondence in a compliant manner. Because it is compliant, the robot is safer for the nearby person or environment, as potential unforeseen collisions will result in lower impact forces. We show the results of motion transfer of squatting from a human to a simulated CoMaN humanoid robot.}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } Motion transfer from a human to a robot implies accurate tracking of desired, demonstrated trajectories. However, direct imitation of joint position trajectories might not result in similar behavior of the robot and the human, because they have different kinematic and dynamic properties, i.e., different embodiment. To avoid the correspondence problem, the demonstrated trajectories need to be somehow adapted. In this paper we go beyond simple imitation, but we show how the torque profiles that should execute the demonstrated position trajectories are being learned in a manner that preserves the correspondence. Thus, position trajectories are modified from the demonstration and, furthermore, the robot executes the motion that preserves correspondence in a compliant manner. Because it is compliant, the robot is safer for the nearby person or environment, as potential unforeseen collisions will result in lower impact forces. We show the results of motion transfer of squatting from a human to a simulated CoMaN humanoid robot. |
Žlajpah, L; Petrič, T Virtual Guides for Redundant Robots Using Admittance Control for Path Tracking Tasks Inproceedings pp. 13–23, Springer International Publishing, Cham, 2019, ISBN: 978-3-030-00232-9. @inproceedings{10.1007/978-3-030-00232-9_2, title = {Virtual Guides for Redundant Robots Using Admittance Control for Path Tracking Tasks}, author = {L. Žlajpah and T. Petrič}, isbn = {978-3-030-00232-9}, year = {2019}, date = {2019-01-01}, pages = {13--23}, publisher = {Springer International Publishing}, address = {Cham}, abstract = {Virtual guides are used in human-robot cooperation to support a human performing manipulation tasks. They can act as guidance constrains to assist the user to move in the preferred direction or along desired path, or as forbidden-region constraint which prevent him to move into restricted region of the robot workspace. In this paper we proposed a novel framework that unifies virtual guides using virtual robot approach, which is represented with the admittance control, where a broad class of virtual guides and constraints can be implemented. The dynamic properties and the constraints of the virtual robot can be defined using three sets of parameters and variables: desired motion variables, dynamic parameters (stiffness, damping and inertia) and dead-zones. To validate the approach we implemented it on a KUKA LWR robot for the Buzz-Wire tasks, where the goal is to move a ring along a curved wire.}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } Virtual guides are used in human-robot cooperation to support a human performing manipulation tasks. They can act as guidance constrains to assist the user to move in the preferred direction or along desired path, or as forbidden-region constraint which prevent him to move into restricted region of the robot workspace. In this paper we proposed a novel framework that unifies virtual guides using virtual robot approach, which is represented with the admittance control, where a broad class of virtual guides and constraints can be implemented. The dynamic properties and the constraints of the virtual robot can be defined using three sets of parameters and variables: desired motion variables, dynamic parameters (stiffness, damping and inertia) and dead-zones. To validate the approach we implemented it on a KUKA LWR robot for the Buzz-Wire tasks, where the goal is to move a ring along a curved wire. |
Petrič, T; Peternel, L; Morimoto, J; Babič, J Assistive Arm-Exoskeleton Control Based on Human Muscular Manipulability Journal Article Frontiers in Neurorobotics, 13 , pp. 30, 2019, ISSN: 1662-5218. @article{10.3389/fnbot.2019.00030, title = {Assistive Arm-Exoskeleton Control Based on Human Muscular Manipulability}, author = {T. Petrič and L. Peternel and J. Morimoto and J. Babič}, doi = {10.3389/fnbot.2019.00030}, issn = {1662-5218}, year = {2019}, date = {2019-01-01}, journal = {Frontiers in Neurorobotics}, volume = {13}, pages = {30}, abstract = {This paper introduces a novel control framework for an arm exoskeleton that takes into account force of the human arm. In contrast to the conventional exoskeleton controllers where the assistance is provided without considering the human arm biomechanical force manipulability properties, we propose a control approach based on the arm muscular manipulability. The proposed control framework essentially reshapes the anisotropic force manipulability into the endpoint force manipulability that is invariant with respect to the direction in the entire workspace of the arm. This allows users of the exoskeleton to perform tasks effectively in the whole range of the workspace, even in areas that are normally unsuitable due to the low force manipulability of the human arm. We evaluated the proposed control framework with real robot experiments where subjects wearing an arm exoskeleton were asked to move a weight between several locations. The results show that the proposed control framework does not affect the normal movement behavior of the users while effectively reduces user effort in the area of low manipulability. Particularly, the proposed approach augments the human arm force manipulability to execute tasks equally well in the entire workspace of the arm.}, keywords = {}, pubstate = {published}, tppubtype = {article} } This paper introduces a novel control framework for an arm exoskeleton that takes into account force of the human arm. In contrast to the conventional exoskeleton controllers where the assistance is provided without considering the human arm biomechanical force manipulability properties, we propose a control approach based on the arm muscular manipulability. The proposed control framework essentially reshapes the anisotropic force manipulability into the endpoint force manipulability that is invariant with respect to the direction in the entire workspace of the arm. This allows users of the exoskeleton to perform tasks effectively in the whole range of the workspace, even in areas that are normally unsuitable due to the low force manipulability of the human arm. We evaluated the proposed control framework with real robot experiments where subjects wearing an arm exoskeleton were asked to move a weight between several locations. The results show that the proposed control framework does not affect the normal movement behavior of the users while effectively reduces user effort in the area of low manipulability. Particularly, the proposed approach augments the human arm force manipulability to execute tasks equally well in the entire workspace of the arm. |
Petrič, T; Žlajpah, L On-line Adaption of Virtual Guides Through Physical Interaction Inproceedings Berns, Karsten; Görges, Daniel (Ed.): Advances in Service and Industrial Robotics, pp. 293–300, Springer International Publishing, Cham, 2019, ISBN: 978-3-030-19648-6. @inproceedings{10.1007/978-3-030-19648-6_34, title = {On-line Adaption of Virtual Guides Through Physical Interaction}, author = {T. Petrič and L. Žlajpah}, editor = {Karsten Berns and Daniel Görges}, isbn = {978-3-030-19648-6}, year = {2019}, date = {2019-01-01}, booktitle = {Advances in Service and Industrial Robotics}, pages = {293--300}, publisher = {Springer International Publishing}, address = {Cham}, abstract = {Virtual guide framework allows efficient learning and control of complex robot behaviors in human-robot interaction scenarios. The framework can help to guide users to move in a predefined direction or prevent them to enter a forbidden-region. As such, the framework also allows efficient modulation of regions by changing of parameters. In this paper, we introduce and evaluate the means of adapting path parameters through physical interaction. The main goal was to introduce an algorithm into a virtual guide framework which can partially modify the path trajectories. The path updates are based on physical interaction and allow human intervention to improve the task performance. This enables to update the path trajectory only where needed and hence, to bypass the need to re-learn the whole trajectory from scratch. Since virtual guides are also active while learning, the required effort from the user is lower compared to the required effort when the user is teaching the robot with kinesthetic guidance. The effectiveness of the proposed algorithm has been demonstrated with simulation results and experiments on a KUKA LWR robot.}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } Virtual guide framework allows efficient learning and control of complex robot behaviors in human-robot interaction scenarios. The framework can help to guide users to move in a predefined direction or prevent them to enter a forbidden-region. As such, the framework also allows efficient modulation of regions by changing of parameters. In this paper, we introduce and evaluate the means of adapting path parameters through physical interaction. The main goal was to introduce an algorithm into a virtual guide framework which can partially modify the path trajectories. The path updates are based on physical interaction and allow human intervention to improve the task performance. This enables to update the path trajectory only where needed and hence, to bypass the need to re-learn the whole trajectory from scratch. Since virtual guides are also active while learning, the required effort from the user is lower compared to the required effort when the user is teaching the robot with kinesthetic guidance. The effectiveness of the proposed algorithm has been demonstrated with simulation results and experiments on a KUKA LWR robot. |
Lukić, B; Petrič, T; Žlajpah, L; Jovanović, K KUKA LWR Robot Cartesian Stiffness Control Based on Kinematic Redundancy Inproceedings Berns, Karsten; Görges, Daniel (Ed.): Advances in Service and Industrial Robotics, pp. 310–318, Springer International Publishing, Cham, 2019, ISBN: 978-3-030-19648-6. @inproceedings{10.1007/978-3-030-19648-6_34b, title = {KUKA LWR Robot Cartesian Stiffness Control Based on Kinematic Redundancy}, author = {B Lukić and T Petrič and L Žlajpah and K Jovanović}, editor = {Karsten Berns and Daniel Görges}, isbn = {978-3-030-19648-6}, year = {2019}, date = {2019-01-01}, booktitle = {Advances in Service and Industrial Robotics}, pages = {310--318}, publisher = {Springer International Publishing}, address = {Cham}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } |
Žlajpah, L; Petrič, T Bounded Self-motion of Functional Redundant Robots Inproceedings Berns, Karsten; Görges, Daniel (Ed.): Advances in Service and Industrial Robotics, pp. 285–292, Springer International Publishing, Cham, 2019, ISBN: 978-3-030-19648-6. @inproceedings{Zlajpah2019_raad, title = {Bounded Self-motion of Functional Redundant Robots}, author = {L Žlajpah and T Petrič}, editor = {Karsten Berns and Daniel Görges}, isbn = {978-3-030-19648-6}, year = {2019}, date = {2019-01-01}, booktitle = {Advances in Service and Industrial Robotics}, pages = {285--292}, publisher = {Springer International Publishing}, address = {Cham}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } |
Jovanović, Kosta; Petrič, Tadej; Tsuji, Toshiaki; Oddo, Calogero Maria Editorial: Human-Like Advances in Robotics: Motion, Actuation, Sensing, Cognition and Control Journal Article Frontiers in Neurorobotics, 13 , pp. 85, 2019, ISSN: 1662-5218. @article{10.3389/fnbot.2019.00085, title = {Editorial: Human-Like Advances in Robotics: Motion, Actuation, Sensing, Cognition and Control}, author = {Kosta Jovanović and Tadej Petrič and Toshiaki Tsuji and Calogero Maria Oddo}, url = {https://www.frontiersin.org/article/10.3389/fnbot.2019.00085}, doi = {10.3389/fnbot.2019.00085}, issn = {1662-5218}, year = {2019}, date = {2019-01-01}, journal = {Frontiers in Neurorobotics}, volume = {13}, pages = {85}, keywords = {}, pubstate = {published}, tppubtype = {article} } |
2018 |
Petrič, T; Gams, A; Colasanto, L; Ijspeert, A J; Ude, A Accelerated Sensorimotor Learning of Compliant Movement Primitives Journal Article IEEE Transactions on Robotics, 34 (6), pp. 1636-1642, 2018, ISSN: 1552-3098. @article{8437179, title = {Accelerated Sensorimotor Learning of Compliant Movement Primitives}, author = {T. Petrič and A. Gams and L. Colasanto and A. J. Ijspeert and A. Ude}, doi = {10.1109/TRO.2018.2861921}, issn = {1552-3098}, year = {2018}, date = {2018-12-01}, journal = {IEEE Transactions on Robotics}, volume = {34}, number = {6}, pages = {1636-1642}, abstract = {Autonomous trajectory generation through generalization requires a database of motion, which can be difficult and time consuming to obtain. In this paper, we propose a method for autonomous expansion of a database for the generation of compliant and accurate motion, achieved through the framework of compliant movement primitives (CMPs). These combine task-specific kinematic and corresponding feed-forward dynamic trajectories. The framework allows for generalization and modulation of dynamic behavior. Inspired by human sensorimotor learning abilities, we propose a novel method that can autonomously learn task-specific torque primitives (TPs) associated to given kinematic trajectories, encoded as dynamic movement primitives. The proposed algorithm is completely autonomous, and can be used to rapidly generate and expand the CMP database. Since CMPs are parameterized, statistical generalization can be used to obtain an initial TP estimate of a new CMP. Thereby, the learning rate of new CMPs can be significantly improved. The evaluation of the proposed approach on a Kuka LWR-4 robot performing a peg-in-hole task shows fast TP acquisition and accurate generalization estimates in real-world scenarios.}, keywords = {}, pubstate = {published}, tppubtype = {article} } Autonomous trajectory generation through generalization requires a database of motion, which can be difficult and time consuming to obtain. In this paper, we propose a method for autonomous expansion of a database for the generation of compliant and accurate motion, achieved through the framework of compliant movement primitives (CMPs). These combine task-specific kinematic and corresponding feed-forward dynamic trajectories. The framework allows for generalization and modulation of dynamic behavior. Inspired by human sensorimotor learning abilities, we propose a novel method that can autonomously learn task-specific torque primitives (TPs) associated to given kinematic trajectories, encoded as dynamic movement primitives. The proposed algorithm is completely autonomous, and can be used to rapidly generate and expand the CMP database. Since CMPs are parameterized, statistical generalization can be used to obtain an initial TP estimate of a new CMP. Thereby, the learning rate of new CMPs can be significantly improved. The evaluation of the proposed approach on a Kuka LWR-4 robot performing a peg-in-hole task shows fast TP acquisition and accurate generalization estimates in real-world scenarios. |
Peternel, L; Petrič, T; Babič, J Robotic assembly solution by human-in-the-loop teaching method based on real-time stiffness modulation Journal Article Autonomous Robots, 42 (1), pp. 1–17, 2018, ISSN: 1573-7527. @article{Peternel2018, title = {Robotic assembly solution by human-in-the-loop teaching method based on real-time stiffness modulation}, author = {L. Peternel and T. Petrič and J. Babič}, doi = {10.1007/s10514-017-9635-z}, issn = {1573-7527}, year = {2018}, date = {2018-01-01}, journal = {Autonomous Robots}, volume = {42}, number = {1}, pages = {1--17}, abstract = {We propose a novel human-in-the-loop approach for teaching robots how to solve assembly tasks in unpredictable and unstructured environments. In the proposed method the human sensorimotor system is integrated into the robot control loop though a teleoperation setup. The approach combines a 3-DoF end-effector force feedback with an interface for modulation of the robot end-effector stiffness. When operating in unpredictable and unstructured environments, modulation of limb impedance is essential in terms of successful task execution, stability and safety. We developed a novel hand-held stiffness control interface that is controlled by the motion of the human finger. A teaching approach was then used to achieve autonomous robot operation. In the experiments, we analysed and solved two part-assembly tasks: sliding a bolt fitting inside a groove and driving a self-tapping screw into a material of unknown properties. We experimentally compared the proposed method to complementary robot learning methods and analysed the potential benefits of direct stiffness modulation in the force-feedback teleoperation.}, keywords = {}, pubstate = {published}, tppubtype = {article} } We propose a novel human-in-the-loop approach for teaching robots how to solve assembly tasks in unpredictable and unstructured environments. In the proposed method the human sensorimotor system is integrated into the robot control loop though a teleoperation setup. The approach combines a 3-DoF end-effector force feedback with an interface for modulation of the robot end-effector stiffness. When operating in unpredictable and unstructured environments, modulation of limb impedance is essential in terms of successful task execution, stability and safety. We developed a novel hand-held stiffness control interface that is controlled by the motion of the human finger. A teaching approach was then used to achieve autonomous robot operation. In the experiments, we analysed and solved two part-assembly tasks: sliding a bolt fitting inside a groove and driving a self-tapping screw into a material of unknown properties. We experimentally compared the proposed method to complementary robot learning methods and analysed the potential benefits of direct stiffness modulation in the force-feedback teleoperation. |
Petrič, T; Cevzar, M; Babič, J Shared Control for Human-Robot Cooperative Manipulation Tasks Inproceedings Ferraresi, Carlo; Quaglia, Giuseppe (Ed.): pp. 787–796, Springer International Publishing, Cham, 2018, ISBN: 978-3-319-61276-8. @inproceedings{Petrič2018, title = {Shared Control for Human-Robot Cooperative Manipulation Tasks}, author = {T. Petrič and M. Cevzar and J. Babič}, editor = {Carlo Ferraresi and Giuseppe Quaglia}, isbn = {978-3-319-61276-8}, year = {2018}, date = {2018-01-01}, pages = {787--796}, publisher = {Springer International Publishing}, address = {Cham}, abstract = {In the past decade many studies on human motor control have investigated how humans are moving their arms. In robotics, these studies were usually used as a foundation for human-robot cooperation tasks. Nonetheless, the gap between human motor control and robot control remains challenging. In this paper we investigated, how human proprioceptive abilities could enhance performance of cooperative manipulative tasks, where humans and robots are autonomous agents coupled through physical interaction. In such setups, the robot movements are usually accurate but without the proprioceptive capabilities observed in humans. On the contrary, humans have well developed proprioceptive capabilities, but their movement accuracy is highly dependent on the speed of movement. In this paper we proposed an approach where we exploited the speed-accuracy trade-off model of a human together with the robotic partner. In this way the performance can be improved in a human-robot cooperative setup. The performance was analyzed on a task where a long object, i.e. a pipe, needs to be manipulated into a groove with different tolerances. We tested the accuracy and efficiency of performing the task. The results show that the proposed approach can successfully estimate human behavior and successfully perform the task.}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } In the past decade many studies on human motor control have investigated how humans are moving their arms. In robotics, these studies were usually used as a foundation for human-robot cooperation tasks. Nonetheless, the gap between human motor control and robot control remains challenging. In this paper we investigated, how human proprioceptive abilities could enhance performance of cooperative manipulative tasks, where humans and robots are autonomous agents coupled through physical interaction. In such setups, the robot movements are usually accurate but without the proprioceptive capabilities observed in humans. On the contrary, humans have well developed proprioceptive capabilities, but their movement accuracy is highly dependent on the speed of movement. In this paper we proposed an approach where we exploited the speed-accuracy trade-off model of a human together with the robotic partner. In this way the performance can be improved in a human-robot cooperative setup. The performance was analyzed on a task where a long object, i.e. a pipe, needs to be manipulated into a groove with different tolerances. We tested the accuracy and efficiency of performing the task. The results show that the proposed approach can successfully estimate human behavior and successfully perform the task. |
2017 |
Petrič, T; Cevzar, M; Babič, J Utilizing speed-accuracy trade-off models for human-robot coadaptation during cooperative groove fitting task Inproceedings 2017 IEEE-RAS 17th International Conference on Humanoid Robotics (Humanoids), pp. 107-112, 2017, ISSN: 2164-0580. @inproceedings{Petrič2017, title = {Utilizing speed-accuracy trade-off models for human-robot coadaptation during cooperative groove fitting task}, author = {T. Petrič and M. Cevzar and J. Babič}, doi = {10.1109/HUMANOIDS.2017.8239544}, issn = {2164-0580}, year = {2017}, date = {2017-11-01}, booktitle = {2017 IEEE-RAS 17th International Conference on Humanoid Robotics (Humanoids)}, pages = {107-112}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } |
Goljat, R; Babič, J; Petrič, T; Peternel, L; Morimoto, J Power-augmentation control approach for arm exoskeleton based on human muscular manipulability Inproceedings 2017 IEEE International Conference on Robotics and Automation (ICRA), pp. 5929-5934, 2017. @inproceedings{Goljat2017, title = {Power-augmentation control approach for arm exoskeleton based on human muscular manipulability}, author = {R. Goljat and J. Babič and T. Petrič and L. Peternel and J. Morimoto}, doi = {10.1109/ICRA.2017.7989698}, year = {2017}, date = {2017-05-01}, booktitle = {2017 IEEE International Conference on Robotics and Automation (ICRA)}, pages = {5929-5934}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } |
Petrič, T; Simpson, C S; Ude, A; Ijspeert, A J Hammering Does Not Fit Fitts' Law Journal Article Frontiers in Computational Neuroscience, 11 , pp. 45, 2017, ISSN: 1662-5188. @article{10.3389/fncom.2017.00045, title = {Hammering Does Not Fit Fitts' Law}, author = {T. Petrič and C. S. Simpson and A. Ude and A. J. Ijspeert}, doi = {10.3389/fncom.2017.00045}, issn = {1662-5188}, year = {2017}, date = {2017-01-01}, journal = {Frontiers in Computational Neuroscience}, volume = {11}, pages = {45}, abstract = {While movement is essential to human wellbeing, we are still unable to reproduce the deftness and robustness of human movement in automatons or completely restore function to individuals with many types of motor impairment. To better understand how the human nervous system plans and controls movements, neuromechanists employ simple tasks such as upper extremity reaches and isometric force tasks. However, these simple tasks rarely consider impacts and may not capture aspects of motor control that arise from real-world complexity. Here we compared existing models of motor control with the results of a periodic targeted impact task extended from Bernstein's seminal work: hammering a nail into wood. We recorded impact forces and kinematics from 10 subjects hammering at different frequencies and with hammers with different physical properties (mass and face area). We found few statistical differences in most measures between different types of hammer, demonstrating human robustness to minor changes in dynamics. Because human motor control is thought to obey optimality principles, we also developed a feedforward optimal simulation with a neuromechanically inspired cost function that reproduces the experimental data. However, Fitts' Law, which relates movement time to distance traveled and target size, did not match our experimental data. We therefore propose a new model in which the distance moved is a logarithmic function of the time to move that yields better results (R^2 > 0.99 compared to R^2 > 0.88). These results support the argument that humans control movement in an optimal way, but suggest that Fitts' Law may not generalize to periodic impact tasks.}, keywords = {}, pubstate = {published}, tppubtype = {article} } While movement is essential to human wellbeing, we are still unable to reproduce the deftness and robustness of human movement in automatons or completely restore function to individuals with many types of motor impairment. To better understand how the human nervous system plans and controls movements, neuromechanists employ simple tasks such as upper extremity reaches and isometric force tasks. However, these simple tasks rarely consider impacts and may not capture aspects of motor control that arise from real-world complexity. Here we compared existing models of motor control with the results of a periodic targeted impact task extended from Bernstein's seminal work: hammering a nail into wood. We recorded impact forces and kinematics from 10 subjects hammering at different frequencies and with hammers with different physical properties (mass and face area). We found few statistical differences in most measures between different types of hammer, demonstrating human robustness to minor changes in dynamics. Because human motor control is thought to obey optimality principles, we also developed a feedforward optimal simulation with a neuromechanically inspired cost function that reproduces the experimental data. However, Fitts' Law, which relates movement time to distance traveled and target size, did not match our experimental data. We therefore propose a new model in which the distance moved is a logarithmic function of the time to move that yields better results (R^2 > 0.99 compared to R^2 > 0.88). These results support the argument that humans control movement in an optimal way, but suggest that Fitts' Law may not generalize to periodic impact tasks. |
2016 |
Petrič, T; Goljat, R; Babič, J Cooperative human-robot control based on Fitts' law Inproceedings 2016 IEEE-RAS 16th International Conference on Humanoid Robots (Humanoids), pp. 345-350, 2016, ISSN: 2164-0580. @inproceedings{Petrič2016, title = {Cooperative human-robot control based on Fitts' law}, author = {T. Petrič and R. Goljat and J. Babič}, doi = {10.1109/HUMANOIDS.2016.7803299}, issn = {2164-0580}, year = {2016}, date = {2016-11-01}, booktitle = {2016 IEEE-RAS 16th International Conference on Humanoid Robots (Humanoids)}, pages = {345-350}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } |
Deniša, M; Gams, A; Ude, A; Petrič, T Learning Compliant Movement Primitives Through Demonstration and Statistical Generalization Journal Article IEEE/ASME Transactions on Mechatronics, 21 (5), pp. 2581-2594, 2016, ISSN: 1083-4435. @article{7360201, title = {Learning Compliant Movement Primitives Through Demonstration and Statistical Generalization}, author = {M. Deniša and A. Gams and A. Ude and T. Petrič}, doi = {10.1109/TMECH.2015.2510165}, issn = {1083-4435}, year = {2016}, date = {2016-10-01}, journal = {IEEE/ASME Transactions on Mechatronics}, volume = {21}, number = {5}, pages = {2581-2594}, abstract = {In this paper, we address the problem of simultaneously achieving low trajectory tracking errors and compliant control without using explicit mathematical models of task dynamics. To achieve this goal, we propose a new movement representation called compliant movement primitives (CMPs), which encodes position trajectory and associated torque profiles and can be learned from a single user demonstration. With the proposed control framework, the robot can remain compliant and consequently safe for humans sharing its workspace, even if high trajectory tracking accuracy is required. We developed a statistical learning approach that can use a database of existing CMPs and compute new ones, adapted for novel task variations. The proposed approach was evaluated on a Kuka LWR-4 robot performing 1) a discrete pick-and-place task with objects of varying weight and 2) a periodic handle turning operation. The evaluation of the discrete task showed a 15-fold decrease of the tracking error while exhibiting compliant behavior compared to the standard feedback control approach. It also indicated no significant rise in the tracking error while using generalized primitives computed by the statistical learning method. With respect to unforeseen collisions, the proposed approach resulted in a 75% drop of contact forces compared to standard feedback control. The periodic task demonstrated on-line use of the proposed approach to accomplish a task of handle turning.}, keywords = {}, pubstate = {published}, tppubtype = {article} } In this paper, we address the problem of simultaneously achieving low trajectory tracking errors and compliant control without using explicit mathematical models of task dynamics. To achieve this goal, we propose a new movement representation called compliant movement primitives (CMPs), which encodes position trajectory and associated torque profiles and can be learned from a single user demonstration. With the proposed control framework, the robot can remain compliant and consequently safe for humans sharing its workspace, even if high trajectory tracking accuracy is required. We developed a statistical learning approach that can use a database of existing CMPs and compute new ones, adapted for novel task variations. The proposed approach was evaluated on a Kuka LWR-4 robot performing 1) a discrete pick-and-place task with objects of varying weight and 2) a periodic handle turning operation. The evaluation of the discrete task showed a 15-fold decrease of the tracking error while exhibiting compliant behavior compared to the standard feedback control approach. It also indicated no significant rise in the tracking error while using generalized primitives computed by the statistical learning method. With respect to unforeseen collisions, the proposed approach resulted in a 75% drop of contact forces compared to standard feedback control. The periodic task demonstrated on-line use of the proposed approach to accomplish a task of handle turning. |
Petrič, T; Goljat, R; Babič, J Augmentation of human arm motor control by isotropic force manipulability Inproceedings 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pp. 696-701, 2016, ISSN: 2153-0866. @inproceedings{Petrič2016b, title = {Augmentation of human arm motor control by isotropic force manipulability}, author = {T. Petrič and R. Goljat and J. Babič}, doi = {10.1109/IROS.2016.7759128}, issn = {2153-0866}, year = {2016}, date = {2016-10-01}, booktitle = {2016 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)}, pages = {696-701}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } |
Peternel, L; Noda, T; Petrič, T; Ude, A; Morimoto, J; Babič, J Adaptive Control of Exoskeleton Robots for Periodic Assistive Behaviours Based on EMG Feedback Minimisation Journal Article PLOS ONE, 11 (2), pp. 1-26, 2016. @article{Peternel2016, title = {Adaptive Control of Exoskeleton Robots for Periodic Assistive Behaviours Based on EMG Feedback Minimisation}, author = {L. Peternel and T. Noda and T. Petrič and A. Ude and J. Morimoto and J. Babič}, doi = {10.1371/journal.pone.0148942}, year = {2016}, date = {2016-01-01}, journal = {PLOS ONE}, volume = {11}, number = {2}, pages = {1-26}, publisher = {Public Library of Science}, abstract = {In this paper we propose an exoskeleton control method for adaptive learning of assistive joint torque profiles in periodic tasks. We use human muscle activity as feedback to adapt the assistive joint torque behaviour in a way that the muscle activity is minimised. The user can then relax while the exoskeleton takes over the task execution. If the task is altered and the existing assistive behaviour becomes inadequate, the exoskeleton gradually adapts to the new task execution so that the increased muscle activity caused by the new desired task can be reduced. The advantage of the proposed method is that it does not require biomechanical or dynamical models. Our proposed learning system uses Dynamical Movement Primitives (DMPs) as a trajectory generator and parameters of DMPs are modulated using Locally Weighted Regression. Then, the learning system is combined with adaptive oscillators that determine the phase and frequency of motion according to measured Electromyography (EMG) signals. We tested the method with real robot experiments where subjects wearing an elbow exoskeleton had to move an object of an unknown mass according to a predefined reference motion. We further evaluated the proposed approach on a whole-arm exoskeleton to show that it is able to adaptively derive assistive torques even for multiple-joint motion.}, keywords = {}, pubstate = {published}, tppubtype = {article} } In this paper we propose an exoskeleton control method for adaptive learning of assistive joint torque profiles in periodic tasks. We use human muscle activity as feedback to adapt the assistive joint torque behaviour in a way that the muscle activity is minimised. The user can then relax while the exoskeleton takes over the task execution. If the task is altered and the existing assistive behaviour becomes inadequate, the exoskeleton gradually adapts to the new task execution so that the increased muscle activity caused by the new desired task can be reduced. The advantage of the proposed method is that it does not require biomechanical or dynamical models. Our proposed learning system uses Dynamical Movement Primitives (DMPs) as a trajectory generator and parameters of DMPs are modulated using Locally Weighted Regression. Then, the learning system is combined with adaptive oscillators that determine the phase and frequency of motion according to measured Electromyography (EMG) signals. We tested the method with real robot experiments where subjects wearing an elbow exoskeleton had to move an object of an unknown mass according to a predefined reference motion. We further evaluated the proposed approach on a whole-arm exoskeleton to show that it is able to adaptively derive assistive torques even for multiple-joint motion. |
Gams, A; Petrič, T; Do, M; Nemec, B; Morimoto, J; Asfour, T; Ude, A Adaptation and coaching of periodic motion primitives through physical and visual interaction Journal Article Robotics and Autonomous Systems, 75 , pp. 340 - 351, 2016, ISSN: 0921-8890. @article{GAMS2016340, title = {Adaptation and coaching of periodic motion primitives through physical and visual interaction}, author = {A. Gams and T. Petrič and M. Do and B. Nemec and J. Morimoto and T. Asfour and A. Ude}, doi = {https://doi.org/10.1016/j.robot.2015.09.011}, issn = {0921-8890}, year = {2016}, date = {2016-01-01}, journal = {Robotics and Autonomous Systems}, volume = {75}, pages = {340 - 351}, abstract = {In this paper we propose and evaluate a control system to (1) learn and (2) adapt robot motion for continuous non-rigid contact with the environment. We present the approach in the context of wiping surfaces with robots. Our approach is based on learning by demonstration. First an initial periodic motion, covering the essence of the wiping task, is transferred from a human to a robot. The system extracts and learns one period of motion. Once the user/demonstrator is content with the motion, the robot seeks and establishes contact with a given surface, maintaining a predefined force of contact through force feedback. The shape of the surface is encoded for the complete period of motion, but the robot can adapt to a different surface, perturbations or obstacles. The novelty stems from the fact that the feedforward component is learned and encoded in a dynamic movement primitive. By using the feedforward component, the feedback component is greatly reduced if not completely canceled. Finally, if the user is not satisfied with the periodic pattern, he/she can change parts of motion through predefined gestures or through physical contact in a manner of a tutor or a coach. The complete system thus allows not only a transfer of motion, but a transfer of motion with matching correspondences, i.e. wiping motion is constrained to maintain physical contact with the surface to be wiped. The interface for both learning and adaptation is simple and intuitive and allows for fast and reliable knowledge transfer to the robot. Simulated and real world results in the application domain of wiping a surface are presented on three different robotic platforms. Results of the three robotic platforms, namely a 7 degree-of-freedom Kuka LWR-4 robot, the ARMAR-IIIa humanoid platform and the Sarcos CB-i humanoid robot, depict different methods of adaptation to the environment and coaching.}, keywords = {}, pubstate = {published}, tppubtype = {article} } In this paper we propose and evaluate a control system to (1) learn and (2) adapt robot motion for continuous non-rigid contact with the environment. We present the approach in the context of wiping surfaces with robots. Our approach is based on learning by demonstration. First an initial periodic motion, covering the essence of the wiping task, is transferred from a human to a robot. The system extracts and learns one period of motion. Once the user/demonstrator is content with the motion, the robot seeks and establishes contact with a given surface, maintaining a predefined force of contact through force feedback. The shape of the surface is encoded for the complete period of motion, but the robot can adapt to a different surface, perturbations or obstacles. The novelty stems from the fact that the feedforward component is learned and encoded in a dynamic movement primitive. By using the feedforward component, the feedback component is greatly reduced if not completely canceled. Finally, if the user is not satisfied with the periodic pattern, he/she can change parts of motion through predefined gestures or through physical contact in a manner of a tutor or a coach. The complete system thus allows not only a transfer of motion, but a transfer of motion with matching correspondences, i.e. wiping motion is constrained to maintain physical contact with the surface to be wiped. The interface for both learning and adaptation is simple and intuitive and allows for fast and reliable knowledge transfer to the robot. Simulated and real world results in the application domain of wiping a surface are presented on three different robotic platforms. Results of the three robotic platforms, namely a 7 degree-of-freedom Kuka LWR-4 robot, the ARMAR-IIIa humanoid platform and the Sarcos CB-i humanoid robot, depict different methods of adaptation to the environment and coaching. |
Petrič, T; Ude, A; Ijspeert, A J Autonomous Learning of Internal Dynamic Models for Reaching Tasks Inproceedings Borangiu, Theodor (Ed.): pp. 439–447, Springer International Publishing, Cham, 2016, ISBN: 978-3-319-21290-6. @inproceedings{10.1007/978-3-319-21290-6_44, title = {Autonomous Learning of Internal Dynamic Models for Reaching Tasks}, author = {T. Petrič and A. Ude and A. J. Ijspeert}, editor = {Theodor Borangiu}, isbn = {978-3-319-21290-6}, year = {2016}, date = {2016-01-01}, pages = {439--447}, publisher = {Springer International Publishing}, address = {Cham}, abstract = {The paper addresses the problem of learning internal task-specific dynamic models for a reaching task. Using task-specific dynamic models is crucial for achieving both high tracking accuracy and compliant behaviour, which improves safety concerns while working in unstructured environment or with humans. The proposed approach uses programming by demonstration to learn new task-related movements encoded as Compliant Movement Primitives (CMPs). CMPs are a combination of position trajectories encoded in a form of Dynamic Movement Primitives (DMPs) and corresponding task-specific Torque Primitives (TPs) encoded as a linear combination of kernel functions. Unlike the DMPs, TPs cannot be directly acquired from user demonstrations. Inspired by the human sensorimotor learning ability we propose a novel method which autonomously learns task-specific TPs, based on a given kinematic trajectory in DMPs.}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } The paper addresses the problem of learning internal task-specific dynamic models for a reaching task. Using task-specific dynamic models is crucial for achieving both high tracking accuracy and compliant behaviour, which improves safety concerns while working in unstructured environment or with humans. The proposed approach uses programming by demonstration to learn new task-related movements encoded as Compliant Movement Primitives (CMPs). CMPs are a combination of position trajectories encoded in a form of Dynamic Movement Primitives (DMPs) and corresponding task-specific Torque Primitives (TPs) encoded as a linear combination of kernel functions. Unlike the DMPs, TPs cannot be directly acquired from user demonstrations. Inspired by the human sensorimotor learning ability we propose a novel method which autonomously learns task-specific TPs, based on a given kinematic trajectory in DMPs. |
2015 |
Petrič, T; Colasanto, L; Gams, A; Ude, A; Ijspeert, A J Bio-inspired learning and database expansion of Compliant Movement Primitives Inproceedings 2015 IEEE-RAS 15th International Conference on Humanoid Robots (Humanoids), pp. 346-351, 2015. @inproceedings{7363573, title = {Bio-inspired learning and database expansion of Compliant Movement Primitives}, author = {T. Petrič and L. Colasanto and A. Gams and A. Ude and A. J. Ijspeert}, doi = {10.1109/HUMANOIDS.2015.7363573}, year = {2015}, date = {2015-11-01}, booktitle = {2015 IEEE-RAS 15th International Conference on Humanoid Robots (Humanoids)}, pages = {346-351}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } |
Nemec, B; Petrič, T; Ude, A Force adaptation with recursive regression Iterative Learning Controller Inproceedings 2015 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pp. 2835-2841, 2015. @inproceedings{Nemec2015, title = {Force adaptation with recursive regression Iterative Learning Controller}, author = {B. Nemec and T. Petrič and A. Ude}, doi = {10.1109/IROS.2015.7353767}, year = {2015}, date = {2015-09-01}, booktitle = {2015 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)}, pages = {2835-2841}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } |
Deniša, M; Gams, A; Ude, A; Petrič, T Generalization of discrete Compliant Movement Primitives Inproceedings 2015 International Conference on Advanced Robotics (ICAR), pp. 565-572, 2015. @inproceedings{Deniša2015, title = {Generalization of discrete Compliant Movement Primitives}, author = {M. Deniša and A. Gams and A. Ude and T. Petrič}, doi = {10.1109/ICAR.2015.7251512}, year = {2015}, date = {2015-07-01}, booktitle = {2015 International Conference on Advanced Robotics (ICAR)}, pages = {565-572}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } |
Peternel, L; Petrič, T; Babič, J Human-in-the-loop approach for teaching robot assembly tasks using impedance control interface Inproceedings 2015 IEEE International Conference on Robotics and Automation (ICRA), pp. 1497-1502, 2015, ISSN: 1050-4729. @inproceedings{Peternel2015, title = {Human-in-the-loop approach for teaching robot assembly tasks using impedance control interface}, author = {L. Peternel and T. Petrič and J. Babič}, doi = {10.1109/ICRA.2015.7139387}, issn = {1050-4729}, year = {2015}, date = {2015-05-01}, booktitle = {2015 IEEE International Conference on Robotics and Automation (ICRA)}, pages = {1497-1502}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } |
Petrič, T; Gams, A; Likar, N; Žlajpah, L Obstacle Avoidance with Industrial Robots Book Chapter Carbone, Giuseppe; Gomez-Bravo, Fernando (Ed.): Motion and Operation Planning of Robotic Systems: Background and Practical Approaches, pp. 113–145, Springer International Publishing, Cham, 2015, ISBN: 978-3-319-14705-5. @inbook{Petrič2015, title = {Obstacle Avoidance with Industrial Robots}, author = {T. Petrič and A. Gams and N. Likar and L. Žlajpah}, editor = {Giuseppe Carbone and Fernando Gomez-Bravo}, doi = {10.1007/978-3-319-14705-5_5}, isbn = {978-3-319-14705-5}, year = {2015}, date = {2015-01-01}, booktitle = {Motion and Operation Planning of Robotic Systems: Background and Practical Approaches}, pages = {113--145}, publisher = {Springer International Publishing}, address = {Cham}, keywords = {}, pubstate = {published}, tppubtype = {inbook} } |
2014 |
Gams, A; Petric, T; Nemec, B; Ude, A Learning and adaptation of periodic motion primitives based on force feedback and human coaching interaction Inproceedings 2014 IEEE-RAS International Conference on Humanoid Robots, pp. 166-171, 2014, ISSN: 2164-0572. @inproceedings{Gams2014, title = {Learning and adaptation of periodic motion primitives based on force feedback and human coaching interaction}, author = {A. Gams and T. Petric and B. Nemec and A. Ude}, doi = {10.1109/HUMANOIDS.2014.7041354}, issn = {2164-0572}, year = {2014}, date = {2014-11-01}, booktitle = {2014 IEEE-RAS International Conference on Humanoid Robots}, pages = {166-171}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } |
Petrič, T; Gams, A; Žlajpah, L; Ude, A Online learning of task-specific dynamics for periodic tasks Inproceedings 2014 IEEE/RSJ International Conference on Intelligent Robots and Systems, pp. 1790-1795, 2014, ISSN: 2153-0858. @inproceedings{Petrič2014, title = {Online learning of task-specific dynamics for periodic tasks}, author = {T. Petrič and A. Gams and L. Žlajpah and A. Ude}, doi = {10.1109/IROS.2014.6942797}, issn = {2153-0858}, year = {2014}, date = {2014-09-01}, booktitle = {2014 IEEE/RSJ International Conference on Intelligent Robots and Systems}, pages = {1790-1795}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } |
Petrič, T; Gams, A; Žlajpah, L; Ude, A; Morimoto, J Online approach for altering robot behaviors based on human in the loop coaching gestures Inproceedings 2014 IEEE International Conference on Robotics and Automation (ICRA), pp. 4770-4776, 2014, ISSN: 1050-4729. @inproceedings{Petrič2014b, title = {Online approach for altering robot behaviors based on human in the loop coaching gestures}, author = {T. Petrič and A. Gams and L. Žlajpah and A. Ude and J. Morimoto}, doi = {10.1109/ICRA.2014.6907557}, issn = {1050-4729}, year = {2014}, date = {2014-05-01}, booktitle = {2014 IEEE International Conference on Robotics and Automation (ICRA)}, pages = {4770-4776}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } |
Ude, A; Nemec, B; Petrić, T; Morimoto, J Orientation in Cartesian space dynamic movement primitives Inproceedings 2014 IEEE International Conference on Robotics and Automation (ICRA), pp. 2997-3004, 2014, ISSN: 1050-4729. @inproceedings{Ude2014, title = {Orientation in Cartesian space dynamic movement primitives}, author = {A. Ude and B. Nemec and T. Petrić and J. Morimoto}, doi = {10.1109/ICRA.2014.6907291}, issn = {1050-4729}, year = {2014}, date = {2014-05-01}, booktitle = {2014 IEEE International Conference on Robotics and Automation (ICRA)}, pages = {2997-3004}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } |